Data modeling
Problem
A US-based convenience store giant wanted to bring the preference data of customer onto the Responsys platform. The daily data feed is about 45 million with a one-to-many relationship. The marketing platform was unable to handle a huge volume of data for segmentation as it had to scale the whole table for result-set.
Solution
- Lister conducted detailed data discovery sessions with the client data team to understand pain points.
- We revamped the data model to handle a daily preference feed of 45 million
- Lister set up an internal job to make sure only required data is filtered out on daily basis and loaded into temporary table
- The temporary table is truncated and loaded on daily basis for optimized segmentation instead of the whole raw table
Outcomes
- Lister's data model was able to handle 45 million customer preferences information of daily data without any issues.
- The result set for segmentation was now generated in seconds
- No campaign launch was delayed due to effective data model and segmentation
